An effort towards further internationalizing the learning sciences provides new exciting directions for my research. In particular, this effort provides a valuable opportunity to explore questions related to the connection between conversation and learning from a cultural standpoint.
In recent years there has been a growing awareness that research in the Learning Sciences community could be significantly strengthened by intensifying the extent to which it draws in researchers from underrepresented regions and begins to investigate learning sciences questions in a controlled way across international lines. While the methodologies the mainstream learning sciences communities have developed and applied in studies that have primarily taken place in Europe or North America have served to build up a substantial body of knowledge that we can be proud of, we understand as a community that we stand to gain tremendous insights from diversifying our target student populations to include regions such as East Asia, South Asia (including India), as well as Africa. In response to this, in collaboration with Raj Reddy, Matt Kam, Bhiksha Raj, and the Pittsburgh Science of Learning Center I am taking a leading role in an effort to start what I am referring to as LearnLab India, as an outgrowth of the PSLC LearnLab model. From a research perspective, LearnLab India is meant to provide the opportunity to investigate learning sciences issues in a rigorous way from an international perspective.
The PSLC LearnLab model has been successful in over 50 schools in the United States. Its paradigm for classroom studies that emphasize both internal and external validity brings state-of-the-art intelligent tutoring technology into realistic learning contexts, including technologies such as Cognitive Tutors (Koedinger et al., 1997), tutorial dialogue technology (Rosé et al., 2001; Rosé & VanLehn, 2005; Kumar & Rosé, 2010), and Collaborative Learning environments that trigger context sensitive support (Kumar et al., 2007; Chaudhuri et al., 2009). Furthermore, it brings the ability to study human learning in a rigorous way through analysis of logged data related to student interactions with and through the technology on a moment by moment basis through application of datamining, text mining, and speech mining technology, which my group has also taken an active role in (Prata et al., 2009; Gweon et al., 2009; Rosé et al., 2008). LearnLab India is a partnership between PSLC and the larger School of Computer Science community at Carnegie Mellon University with the Indian Institute for Information Technologies in Hyderabad (IIIT-H). LearnLab India is housed at the three campuses of Rajiv Gandhi University of Knowledge Technologies (RGUKT), which is an educational outreach by IIIT-H to the rural youth of Andhra Pradesh, India. The goal of LearnLab India is to help RGUKT students, most of whom attended non-English-medium schools prior to college, to succeed in English-medium university coursework at RGUKT.
Building both on a foundation being laid within a collaborative project with Sue Fussell at Cornell relating to culture and collaborative conversation as well as work within the Social and Communicative Factors in Learning thrust of the Pittsburgh Science of Learning center where Lauren Resnick and I are exploring properties of classroom discussion facilitation that lead to learning, in this context we can explore how those facilitation strategies must be adapted for use in a different cultural context. In December 2009 I ran a small pilot investigation at RGUKT where I acted as facilitator in two classroom discussions using strategies that have been developed and refined, and successfully evaluated in the United States. These sessions were video recorded, and analysis is currently under way. Early results point towards differences in expectations related to classroom participation practices and roles between Indian classrooms and American classrooms that will need to play a major role in the adaptation of these facilitation techniques.
As an outreach in connection with this effort, the PSLC offered a small amount of sponsorship funds to help start an Internship Program in Technology Supported Education , with the goal of building collaborations related both to teaching and research between institutions of higher learning in India and top ranking universities in the United States, beginning with but not limited to Carnegie Mellon University. That sponsorship helped fund a Winter School in Hyderbad, India taught by me, Bhiksha Raj, and Matt Kam in which 130 undergraduates from around India came together for an intensive two-week program to learn about research in the Learning Sciences and participate in short-term projects related to the LearnLab India effort. Through sponsorship from Microsoft Research India, the Pittsburgh Science of Learning Center, and Nokia, we were able to partly or fully sponsor 9 interns to spend Summer of 2010 in Pittsburgh working with us and other faculty in the School of Computer Science.
One important international collaboration that has already been formed is a partnership with Dr. Vasudeva Varma from IIIT Hyderbad, who I hosted for a short sabbatical at the Language Technologies Institute in Spring of 2010. Through these weeks of working together, we developed a plan for joining forces between the Pittsburgh LearnLab India team and his Hyderbad group working on a Nokia funded Smart Classroom project also at RGUKT. In partnership with RGUKT administrator Praveen Garmiella, I am supervising an effort to author and deploy interactive math practice and assessment technology using tools developed locally through the Assistments project. He sees this as a precursor to the dissertation research related to self-directed learning that he plans to begin in the coming year, with Raj Reddy as his advisor, and with me serving on his committee as an external member. Another effort is under way to deploy technology for English as a second language, beginning with HCII PhD student Ruth Wylie’s article tutor, with plans to expand into dialogue based conversational practice environments building on Rohit Kumar’s pre-dissertation work. These are all infrastructure building efforts to enable future large scale, controlled, international comparative education studies.
The RGUKT campuses have an annual intake of 6,000 students. The magnitude of the size of the student population available in the research partnership with RGUKT would provide a tremendous potential for very large scale, highly controlled classroom studies, which would provide invaluable insights to inform continued research on the effective use of language technologies to support instruction. This partnership would greatly accelerate the effort to develop highly effective computer supported instruction because of that magnitude. It would also allow fine grained and finely controlled international comparison studies such as has never before been possible, which holds the potential to transform the field of international comparative education research. This LearnLab India effort will be featured at a symposium on Internationalizing the Learning Sciences, which will be at the International Conference of the Learning Sciences in Summer 2010 (Rosé & Kam, in press).
We have already begun conducting large scale research studies in connection with this effort. Recently we completed a controlled study in information seeking behavior with approximately 2.000 students at RGUKT, which has been submitted for journal publication (Gupta & Rosé, Under Review). This information literacy related research effort connects with a curriculum development effort I am engaged with along with Jean Alexander from our university library system, the Eberly Center for Teaching Excellence, and the Open Learning Initiative, to develop an information literacy unit for the Computing@Carnegie Mellon course, which all CMU Freshman take. Eventually that course unit will provide the opportunity to collect a large amount of comparison data to further fuel this research. Results from this study also provided the foundation for a collaborative CDI proposal submitted by me, Matt Kam, and Stephan Vogel.